Ten tracks. 204 lessons. One path.
Start at Foundations and work down, or go straight to the model you already use. Everything is visible from the start — no drip feed, no unlocking, no waiting a week for the next module.
Foundations
What AI actually is and how it got here. The words you need. How it fails, what it costs, what never goes in. Then the habits: interrogating an answer, getting a real critique, second opinions, grounded research, a prompt library and a weekly spend review.
Claude
The deepest track. Web, Cowork and Code. Projects, MCP, skills, hooks, subagents, memory and long context. Claude Code properly — locking it down, plan mode, worktrees, cost control and the failure loops. Then the API: streaming, tool use end to end, structured outputs, files and vision, retries and batches.
OpenAI
ChatGPT and the platform behind it. Every plan and what it really unlocks. Projects, memory, canvas, deep research, voice. Then the Responses API: conversation state, background mode, function calling, structured outputs, hosted file search, Batch and caching with worked numbers.
Grok
xAI's models and the real-time angle: server-side web and X search, in the apps and in code. Imagine for images and video. The API in depth, and the three values that repoint an existing client at a second vendor.
Gemini
Google's models at full depth, then Gemini across Google — Gems, memory, Deep Research, Canvas, Workspace, Android, Chrome and NotebookLM. Then the builder's layer: the SDK, thinking levels, JSON schemas, the Files API with real token maths, search grounding, caching and batch.
Open source and self-hosted
Run models on your own machine. The hardware maths, the tools, one lesson each on the main open families, then the harder calls: buying or renting a GPU, fine-tuning, licences, and a fully private local agent.
Images and video
Every generator taken one at a time — setup, real prices, what each is genuinely best at, its prompting quirks, editing and references. Then video properly: storyboard, first frames, takes, upscaling, the cut, and the rights and disclosure rules.
Voice
Text to speech, cloning with consent, the main tools one by one, open-source voice on your own machine, transcription you can trust, realtime agents, dubbing and captions that meet the standard. With what each one costs per minute.
Building things
Where it all combines. The agent loop and how to bound it. Workers, Postgres, row-level security, accounts, payments, environments and secrets. Retrieval, embeddings, grounding and citations, evals, caching. Then the channels: a web widget, SMS, email, Slack, and a scheduled job with a human approval gate.
Getting paid
The bridge between being able to build and being paid to build. What businesses buy over and over and what each is worth. Pricing, deposits and staged payments. Finding the first clients with no employment history, no brand and no network. The discovery call, the proposal, the clauses that protect you, and the handover.
Read it, build it, prove it.
-
Read the lesson
1,400 to 2,200 words that go to the mechanism, with every fact sourced to the maker's own documentation and the date it was checked. Highlight any sentence and ask about that exact line.
-
Do the exercise
One step of one continuous build, ten to thirty minutes, ending in a named file that still exists tomorrow. It works whether or not you tell anyone what you are building.
-
Take the test
Get one wrong and it explains the exact thing you missed, then writes a new question on the same idea. Your skill map fills in per concept and per model, so you can see what is actually solid.
-
Run it in the playground
Every lesson has a live playground. Run your prompt against two vendors at once, see both answers, and see what each one cost.
-
Watch it happen
Each lesson is also filmed. The video sits on the lesson page and on YouTube, including the parts where it goes wrong.